Warpage prediction for unidirectional reinforced thermoplastic extruded profiles

• Main Authors: Lippens, Willem, Ivens, Jan, Desplentere, Frederik
• Format: Conference Proceeding
• Language: English
• Subjects: Continuous extrusion; Continuous production; Cross-sections; Error analysis; Extrusion dies; Extrusion rate; Finite element method; Mechanical properties; Polymers; Reinforcing fibers; Reinforcing materials; Residual stress; Rigid PVC; Temperature dependence; Thermal conductivity; Thermal expansion; Thermal simulation; Thermodynamic properties; Warpage; Water baths
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0204643

Thermoplastic over-extrusion is a continuous production process in which reinforcing UD-fibers are preheated and guided at constant line speed into an extrusion die. Here the fibers are surrounded by molten polymer. After the extrusion die, the composite is cooled by a combination of calibration units and water baths. At the end of the process line, the parts are cut to length. Products made with this process have a (or multiple) reinforced zone(s) in their cross-section, allowing for improved mechanical performance with a minimum of reinforcing materials when strategically located. However, due to the mismatch in coefficient of thermal expansion between the reinforcing fibers and the surrounding polymer, excessive warpage and residual stresses after cooldown are often present in the final products. Reducing these unwanted effects is currently done by a costly and lengthy trial-and-error approach. This study presents a multiphysics (thermal-structural) simulation laying the basis to predict warpage and residual stresses in an over-extruded product with complex cross-section using UD-GF/PET commingled yarns over-extruded with rigid PVC. It makes use of experimentally determined temperature-dependent mechanical properties as well as temperature-dependent thermal properties (specific heat capacity, thermal conductivity). The finite element model is compared to experimental data in order to check the validity. The thermal simulation has a maximum deviation of less than 15% when compared to experimentally determined temperatures. A strong overestimation is observed regarding the warpage prediction when compared to experimental data, which has been attributed to the simulated bonding conditions between the components.